Using the Estimation Tool

Use the Estimation Tool to model workload scenarios and estimate suitable Budget CPU and Budget Memory values for an execution profile.

For background information about what the tool estimates, how to interpret Budget, Simulated, and Max, and the tool’s limitations, see Estimation Tool Overview.

Before you start

Goal

• Use the Estimation Tool to model workload scenarios and estimate suitable Budget CPU and Budget Memory values for an execution profile.

Prerequisites

• A configured Altair SLC Hub server.
• An existing execution profile.
• Altair SLC Hub configuration privileges for the required execution profile.

Open the Estimation Tool

1. Navigate to Hub Administration in the portal.
2. Select Execution Profiles.
3. Open the execution profile you want to edit.
4. Select the Resources tab.
5. Click Open Estimation Tool.

The Estimation Tool opens in a dialog.

Review the current budget

When the tool opens, it loads the current Budget CPU and Budget Memory values from the execution profile.

These values are shown in the Budgeted Resources (per Workload) section.

You can edit these values directly in the tool.

Select worker nodes

In the Maximum Resources section, choose which worker nodes contribute to the available CPU and memory for the scenario.

By default, all available worker nodes are selected.

Selecting or clearing worker nodes changes the total available capacity shown in the tool.

Note

The selected worker nodes are used only for the scenario being modeled in the tool. They do not match constraints configured elsewhere in the execution profile.

Configure the workload simulation

In the Workloads Simulation section, configure how many concurrent workloads of each size to simulate.

Use the workload counts to model the expected mix of workloads in your environment. For example, you can test whether the system can handle mostly typical workloads, many smaller workloads, or a smaller number of larger workloads.

The workload sizes are defined relative to the typical workload you expect:

• Medium represents the average or expected workload size. Configure its CPU and memory values to reflect the size you expect most workloads to be.
• Small represents workloads that require fewer resources than the expected workload size. Use this to explore scenarios with many smaller workloads and to see how that affects resource usage.
• Large represents workloads that require more resources than the expected workload size. Use this to explore scenarios with larger workloads and to assess whether the available resources and configured budget can still support them.

By adjusting the number of Small, Medium, and Large workloads, you can test different workload mixes and see how they affect estimated resource usage, queuing, and warnings.

The tool updates the Usage charts and warnings as you change the workload counts.

Adjust workload sizes

The Average Size values define the estimated CPU and memory usage for Small, Medium, and Large workloads.

When the tool opens, these values are based on the current budget:

• Small starts at half the current budget
• Medium starts at the current budget
• Large starts at double the current budget

Use the Scale All Average Sizes Up and Scale All Average Sizes Down buttons to make all simulated workloads larger or smaller.

Use this to explore more extreme scenarios, such as cases where workloads are much larger or smaller than expected, and to see how that affects resource usage.

Review the charts

The Usage charts compare:

• CPU Used and CPU Budget
• Memory Used and Memory Budget
• the maximum available CPU and memory from the selected worker nodes

Use the charts to understand the current scenario:

• If Simulated is close to or above Max, the scenario may exhaust available capacity.
• If Budget exceeds Max, some workloads may queue.
• If Simulated is greater than Budget, workloads may consume more resources than expected.
• If Budget is much greater than Simulated, the budget may be more generous than necessary.

Review warnings

The Warnings section highlights important conditions in the scenario.

Typical warnings include:

• budgeted resource demand exceeds available capacity
• workloads may fail due to out-of-memory errors

Review these warnings before applying any values from the tool.

Fit the budget to the simulated workloads

Click Fit To Workloads to update the Budget CPU and Budget Memory values so they better match the current workload simulation.

Use this when:

• the simulated workloads represent the scenario you expect
• you want the budget to better reflect estimated workload usage

After using Fit To Workloads, review the charts and warnings again.

Apply or discard changes

At the bottom of the dialog, choose one of the following actions:

• Accept Budget Values — copies the Budget CPU and Budget Memory values from the tool into the execution profile
• Cancel — closes the tool without applying changes
• Revert — resets the dialog state to the original values loaded when the tool was opened

After accepting tool values, save the execution profile to keep the changes.

Tips

• Start with the current budget values and model the workload mix you expect most often.
• Use worker-node selection to compare different capacity scenarios.
• Pay close attention to memory warnings, especially out-of-memory risk.
• If you are unsure of exact workload sizes, use the tool to explore a range of reasonable scenarios rather than searching for a perfect value.